gpt-image-2-6/economy-text-to-image

计费

按基础成本 5 积分 × 后台媒体倍数向上取整；当前 1.2 倍为 6 积分/张。

调用示例

POST /v1/images/generations

{
  "model": "gpt-image-2-6/economy-text-to-image",
  "prompt": "Create a horizontal 3:2 overhead product photograph of a blood orange ice pop resting diagonally on a pale pink tabletop. A torn cobalt-blue paper sleeve covers the lower third at the bottom left. The exposed ice pop extends toward the upper right, with a thin half-moon slice of blood orange frozen just beneath its surface.\n\nDirect low afternoon sunlight through the ruby-orange ice from the upper left. Let the citrus segments glow and cast a small patch of orange light on the tabletop beside the tip. Keep a trace of frost near the torn opening and a few beads of meltwater along the sunlit edge.\n\nPrint “ModelFold” in large cream block capitals across the broad, upward-facing panel of the sleeve. Keep the full name sharp and readable at thumbnail size. Let the paper crinkle around the torn edge while the printed panel lies flat. The sleeve carries the only text. Use a close crop with a little breathing room around the product and its shadow.",
  "aspect_ratio": "auto",
  "background": "auto",
  "output_format": "png",
  "frame_count": "auto",
  "frame_duration_ms": 200
}

返回 data[].url 和 usage；图生图必须传 input_urls 数组。

示例素材